(2019). Being Active in Online Communications: Firm ... · behaviour by identifying firm engagement...
Transcript of (2019). Being Active in Online Communications: Firm ... · behaviour by identifying firm engagement...
Sheng, J. (2019). Being Active in Online Communications: FirmResponsiveness and Customer Engagement Behaviour. Journal of InteractiveMarketing, 46, 40-51. https://doi.org/10.1016/j.intmar.2018.11.004
Peer reviewed version
License (if available):CC BY-NC-ND
Link to published version (if available):10.1016/j.intmar.2018.11.004
Link to publication record in Explore Bristol ResearchPDF-document
This is the author accepted manuscript (AAM). The final published version (version of record) is available onlinevia Elsevier at https://www.sciencedirect.com/science/article/pii/S1094996818300720 . Please refer to anyapplicable terms of use of the publisher.
University of Bristol - Explore Bristol ResearchGeneral rights
This document is made available in accordance with publisher policies. Please cite only the publishedversion using the reference above. Full terms of use are available:http://www.bristol.ac.uk/pure/about/ebr-terms
1
Being Active in Online Communications: Firm Responsiveness and Customer
Engagement Behaviour
Abstract
This research investigates the behavioural effects of firms’ online activeness in influencing
customer engagement in word-of-mouth communications. Using a large-scale field dataset of
hotel reviews and managerial responses, this study empirically examines firm responsiveness
in relationship to community members’ participation in the online review posting. Novel
findings are reported that response volume and speed are important for effecting firm–
customer interactions. This highlights a firm-leading influence on customers’ word-of-mouth
behaviour by identifying firm engagement as a motivational driver of customer engagement.
It offers implications for researchers and practitioners with regard to social media marketing,
in particular firm engaging in the online communication network by acting in an active and
prompt manner.
Keywords: online managerial responses; customer reviews; customer engagement behaviour;
firm engagement; social media marketing
2
1. Introduction
Supported by Web 2.0 (O’Reilly, 2007), information creation and exchange in social media
are increasingly common in the digital world (Kaplan & Haenlein, 2010). In this participatory
environment, many more customers are willing to participate in the online communications to
share their experiences with firms and other members of the community. Such interactive
activities take place in various forms, including blogging, word-of-mouth communications,
writing reviews, and recommendations (Van Doorn et al., 2010). The massive content
customers write can be a useful information source for firms (Kozinets, 2002) in developing
and improving businesses’ dynamic marketing capabilities (Barrales-Molina et al., 2014).
More importantly, engaged customers become online word-of-mouth advertisers for
businesses, and this imposes a profound impact on the deeper level of firm–customer
relationships and long-term business performance.
The power of sharing is determined by the breadth and depth of customer engagement, which
has become an essential feature of businesses. For example, the volume of user-generated
content (UGC) for a product/service or the number of users in an online brand community
serves as an indicator of a brand’s or a product/service’s popularity (Proserpio & Zervas,
2017). Such popularity may attract wider attention, potentially leading to increased
recognition and sales (Tirunillai & Tellis, 2012), according to the social influence network
theory (Friedkin, 1998). This theory posits that the social influence created by online traffic
and the propagated information in a social network is pervasive in shaping individuals’
attitudes, cognition and behaviour (Iyengar et al., 2011; Kurt et al., 2011). Management
researchers regard such social influence among fellow consumers as one of the primary
factors affecting consumers’ choices of products (e.g., Kurt et al., 2011; Wang et al., 2013),
purchase intentions (e.g., Fang et al., 2013; Zhang et al., 2014), perception (e.g., Cheng & Ho,
3
2015; Lee et al., 2015), and online communication behaviour (e.g., Goes et al., 2014; Sridhar
& Srinivasan, 2012; Zhang et al., 2011). A wider scope of customer engagement can amplify
the crowd’s voice on the internet and hence strengthen the social influence, with a greater
number of customers acting and participating in the network (King et al., 2014).
The substantial influence of customer engagement on business performance calls for
marketers to expand network scale to acquire market knowledge, competing for a portion of
public attention, improving online reputation, retaining and satisfying customers and creating
synergistic effects (Chang et al., 2015). To encourage customers to voice their opinions,
companies are now acting in social media. For instance, many firms initiate and manage fan
pages on social networking sites to breed online brand community. The virtual online
community in the computer-mediated social gathering context fosters customers’ engagement
in the network (Shriver et al., 2013), leading to increased trust and network effects of the
social influences on connected people reciprocally (Fang et al., 2013; Shoham et al., 2017).
In addition to interactions among community members, we also observe purposeful
marketing posts by firms and a growing number of online firm–customer conversations such
as chatting with customers and responding to customers’ online posts on review platforms or
discussion forums. Marketing researchers and practitioners have recognised that social media
are becoming a desired and efficient channel connecting consumer and marketers
(Schniederjans et al., 2013; Hollebeek et al., 2014). It is useful for disseminating information
and engaging customers, through which companies impose influence on customers’
behaviour (Evans, 2010).
Given the business impact of customer engagement and thus the importance of business
strategy to encourage customer engagement behaviour via social media, the key question that
motivates this study is that “To what extent can businesses’ social media effort affect
4
customer engagement behaviour?”. A review of the literature does not give a clear answer.
This question is first concerned with the antecedents of customer engagement, for which
prior studies primarily focus on the customer, firm or context specific factors (Van Doorn et
al., 2010). King et al. (2014) illustrate that in the current body of knowledge of the
antecedents of electronic word-of-mouth (eWOM) participation, there is a need to know how
firms can foster reviews and reviewers. In particular, little is known about whether firm
engagement in social media activities, in other words, the firm–customer interaction, is also a
motivational driver for customer engagement in eWOM. Further, this question is also of high
practical relevance. Research on firms’ strategic use of online social sites is in an early stage
(Goh et al., 2013), with current efforts devoted to debating whether to engage in social media
activities and evaluating their economic value. However, little attention has been paid to the
efficacy of firms’ social media efforts in affecting customer engagement rather than purchase
behaviour (Lamberton & Stephen, 2016). Discussion is also lacking on business online
activeness after strategic regime change (i.e., adoption of social media strategy) and how
such activeness continuously impacts customers’ engagement behaviour.
The aim of this research is to empirically investigate the behavioural value of businesses
being responsive online in stimulating customers’ engagement in the eWOM
communications. Specifically, this paper studies online managerial responses to customer
reviews and whether and how businesses’ responsiveness affects community members’
participation in the online review posting. Using review and response data of 1,024 London
hotels over a 15-year period, this research considers online managerial response
characteristics (e.g., volume, speed, length) and tests these instruments in relation to the
number of customer reviews posted on the online review platform.
5
The findings demonstrate that customers’ engagement in writing online reviews, in addition
to individual specific determinants (see Hennig-Thurau et al., 2004), is also influenced by
businesses’ responsiveness in the online interactions. This adds to our knowledge about
engagement by identifying firm engagement as an additional driver for customer engagement
behaviour. This research also provides insights to the social media marketing literature. By
investigating the implicit intervening process in influencing consumers’ mindsets (Srinivasan
et al., 2010), this research discovers factors that influence online responsiveness and further
review volume. It emphasises business activeness in social media activities and the continuity
and consistency of relevant practices to encourage more customers to engage in online
communications.
The article proceeds as follows. It first presents an overview of relevant literature and
discusses the theoretical basis for hypothesising the influence of managerial responsiveness
on customer engagement. The data and sample selection are described in the next section.
Then the tests of the cross-sectional and longitudinal effects of managerial responses are
presented in the following section. Finally, the paper is concluded with a discussion of the
research and managerial implications.
2. Social Media Marketing and Online Managerial Responses
In response to the social sense of business, marketers are gaining enthusiasm for capitalising
on the social context and social influence for marketing activities (Yadav et al., 2013). Social
media marketing is defined by Felix et al. (2017, p. 123) as “an interdisciplinary and cross-
functional concept that uses social media (often in combination with other communications
channels) to achieve organisational goals by creating value for stakeholders”. Kozinets et al.
(2010, p. 71) describe social media marketing as the “intentional influencing of consumer-to-
consumer communications by professional marketing techniques”. Social media marketing
6
takes many forms, such as initiating fan pages on social networking sites and responding to
customers’ comments on review platforms.
Current literature on social media marketing efforts and effectiveness mainly focuses on
economic outcomes. Previous studies have documented that marketers play a persuasive role
in social media, and marketer-generated content can affect customers’ purchase behaviour
(Goh et al., 2013). A positive association between social media marketing and purchase
intention/expenditure (e.g., Kim & Ko, 2012; Kumar et al., 2016; Gong et al., 2017) results
from the increased marketing capabilities built upon social media resources (Wang & Kim,
2017). Apart from driving revenue generation, such networking strategy is also powerful in
brand management (Gensler et al., 2013). Godey et al. (2016) find that social media
marketing favourably influences brand equity, especially brand awareness and brand image,
as well as customers’ behaviour towards the brand such as loyalty and preference. The
creation and spread of firm-to-consumer social messages effectively enhance brand
awareness, consideration and preference and attract new customers (De Vries et al., 2017).
In addition to exchange-related aspects, behavioural consequences of social media marketing
efforts also are evident, particularly with regard to customers’ engagement in online
communications. Customer engagement is “a behavioural manifestation toward the brand or
firm that goes beyond transactions” (Verhoef et al., 2010, p. 247), and “a multi-dimensional
concept comprising relevant cognitive, emotional, and behavioural dimensions”, varying in
different contexts (Hollebeek et al., 2014, p. 152). Voluntary participation in social media can
be both passive (i.e., reading the content generated by others) and active (i.e., creating
content and sharing opinions) (Ashley & Tuten, 2015), which benefits the business, the brand
and/or customers (Dong & Sivakumar, 2017). Incentives for customer engagement behaviour
are multifaceted, involving customer-, firm- and context-related factors (Van Doorn et al.,
7
2010). The literature documents that firms may be motivational drivers for customer
engagement, mainly stemming from brand characteristics, venue/channel support,
information environment and incentive rewards (see Van Doorn et al., 2010). However, very
few studies have paid attention to the firm–customer conversations and firm-generated
content, and the findings of these studies are mixed (e.g., Kumar et al., 2013; De Vries et al.,
2017). In fact, the social media marketing effort may also play a role in “chang[ing] customer
engagement states—including their levels, intensities, and complexity” (Bolton, 2011, p. 273).
Harmeling et al. (2017) define customer engagement marketing as “a firm’s deliberate effort
to motivate, empower, and measure a customer’s voluntary contribution to the firm’s
marketing functions beyond the core, economic transaction” (p. 317). The objective of this
marketing strategy is to motivate customers to actively participate and contribute to the
marketing activities as “pseudo-marketers” (Harmeling et al., 2017, p. 312). Given the aim is
to motivate customer engagement behaviour, it raises questions of how to motivate and how
effective the strategy is. Nevertheless, little attention has been paid to firms’ engagement in
generating content in social media, and there is a lack of empirical evidence showing the
effectiveness of firm–customer interactions on consumers’ participation behaviour in online
communications.
With respect to the nascent area of managerial responses to customer reviews, a handful of
studies seem to suggest the potential impact of providing managerial responses on review
volume. Ye et al. (2010) explore the impact of managerial responses on the volume of
subsequent customer reviews. It applies a difference-in-difference approach to the customer
review and management response data by matching hotels across two online review
platforms. Comparing the volume of reviews before and after the first response, they find a
positive impact of providing responses on review volume, but such influence diminishes if no
further responses are provided. Using a similar cross-platform setting, Proserpio and Zervas
8
(2017) touch on the impact of management responses on review volume when they discuss
the mechanism for response affecting review ratings. They find an increase in review
volume—especially the number of positive reviews—after hotels start to respond. Chevalier
et al. (2017) present similar findings, demonstrating that managerial responses can stimulate
customers’ reviewing activities, particularly critical reviews. Furthermore, by testing a panel
model, Xie et al. (2016) find that managerial responses can lead to an increase in the volume
of subsequent consumer reviews. They attribute the increased number of consumer word-of-
mouth to the online firm–customer interactions.
3. Conceptual Framework and Hypotheses
The review of previous studies reveals two gaps in the literature. One is the inconclusive
discussion on firm engagement in online communication in relation to customer engagement
behaviour. The second gap is a lack of empirical investigation into firm responsiveness and
its effectiveness in making behavioural effects in the online interactive network. Nevertheless,
examining the efficacy of firms’ social media efforts is important because of the public
nature of managerial responses and thus the potential influence on other consumers and
potential reviewers. 1 Therefore, this research focuses on business responsiveness in the
review context and contends that observing managerial responses can be an additional driver
for customers’ engagement in eWOM activities.
A conceptual framework is proposed to depict the determinants of customer engagement
behaviour (see Figure 1). It is argued that business responsiveness (measured by response
�������������������������1 In this study, the term ‘potential reviewer’ refers to existing customers who have not yet written reviews of
their most recent service experience, regardless of whether they have written reviews before. Customers who
have written multiple reviews for a hotel can be identified as returning customers. They may have strong
preferences for the hotel and are more likely to write reviews, but this does not mean they will write reviews for
the hotel again.
9
volume, speed and length), online popularity (i.e., the number of reviewers), and hotel
characteristics that may affect service quality and customer satisfaction (e.g., star class,
customer rating, chain brand, size, and age) are related to future review volume (i.e., the
number of review posts in a future period). 2 In addition, the effects of business
responsiveness can be potentially moderated by hotel specific factors.
------------------------------
Insert Figure 1 about here
------------------------------
Response volume
First, by responding to customer reviews, firms establish their social media presence in the
online virtual community. This is an indicator of business adoption of social media strategy,
which explicitly signals to customers that firms are willing to listen and interact (Proserpio &
Zervas, 2017). In this case, customers’ inferences about business trustworthiness are
enhanced (Sparks et al., 2016). As a result, observing the firm–customer online interactions
may inspire customers to voice with an expectation of that their opinions will be heard and
responded to by the service provider (Gu & Ye, 2014); in other words, engagement is
potentially strengthened (Higgins & Scholer, 2009). In particular, a higher volume of
business-to-customer conversations presents a clearer behavioural manifestation of firm’s
responsiveness to customers’ opinions. This leads to the following hypothesis:
Hypothesis 1: The volume of online managerial responses is positively associated with the
future volume of customer reviews.
�������������������������2 The projection of the relationship between future review volume and responsiveness and between future
review volume and online review popularity is based on the assumption that customers read the reviews and
responses before deciding to write reviews, either before or after their consumption experience.
10
Response speed
The second factor is the speed of responding, measured by the average days between the date
of the review posts and that of the associated responses. A shorter interval implies a faster
responding speed. Service research has documented that timing and speed of response have a
substantial influence in managing complaints and improving trust (e.g., Davidow, 2003;
Homburg & Fürst, 2007; Sparks et al., 2016). In influencing customer engagement behaviour,
the speed of responding performs a symbolic function (Enz & Grover, 1992), signalling that
the firm is active in embracing and managing customers’ comments. It indicates the firm is
devoting efforts to maintaining an interactive relationship with their customers, which is an
essential element of responsiveness. Moreover, response speed is critical to determine the
position of response posts and hence its visibility and the attention it is able to attract from
review readers (De Vries et al., 2012). Customer reviews are normally displayed on sites in
reverse chronological order, with the most recent appearing first and each page only
displaying a few posts. Many reviews are generated on the website every day, pushing older
comments and the associated responses to later pages and making them less observable than
those on the first few pages (De Vries et al., 2012), given the fact that customers hardly ever
go beyond the first few pages (Pavlou & Dimoka, 2006). To compete with the rapid update of
reviews, quicker responses increase the probability of responses being displayed and visible
on the first few pages (Wang & Chaudhry, 2018) and thus influencing potential reviewers. It
is reasonable to hypothesise:
Hypothesis 2: The speed of responding is positively associated with the future volume of
customer reviews.
11
Response length
In addition, the content of conversations may provide incentives for expanded customer
engagement. The online review-response establishes a communication channel connecting
firms and customers and diffusing information (Felix et al., 2017). Management teams often
acknowledge customers’ word-of-mouth contribution, praise or distress, and promise to
address the raised issues or sometimes offer offline benefits or compensation (Davidow,
2003). On the one hand, potential reviewers may be motivated to take advantage of the online
and cost-effective medium with an expectation of their efforts being acknowledged, problems
being solved, or additional benefits being offered (Gu & Ye, 2014). On the other hand, the
functional and social benefits may also prompt consumers to engage in order to obtain
information, establish an interpersonal relationship with firms, and fulfil a social need
(Homburg et al., 2015; Buechel & Berger, 2018). Furthermore, marketer-generated content
can be viewed as advertisements, and the content of such exogenous word-of-mouth (Godes
& Mayzlin, 2009) with a deliberate attempt to market the product or service may raise public
scrutiny. Customers may react to the authentic or exaggerated information supplied by firms
in the online conversations to a greater extent, leading to a higher propensity to engage and
speak out. The length of responses is considered as an indicator of the informational role of
responses; therefore, it is hypothesised that:
Hypothesis 3: The length of online managerial response is positively associated with the
future volume of customer reviews.
Firm characteristics
Furthermore, hotel specific factors may play a part in influencing the number of customer
reviews and the magnitude of response effects. Responding hotels may be inherently better
12
managed and operated with quality service. These hotel-specific attributes could directly or
indirectly contribute to the attractiveness of hotels in terms of offline guest visiting and
online review writing, resulting in an increased or decreased number of online reviews. To
measure hotels’ managerial ability and service quality, star class (i.e., a hotel’s star class on a
five-star scale), customer ratings (i.e., a hotel’s overall average review rating on a scale of 1
to 5), and chain brand (i.e., a hotel is a chain or an independent hotel) are considered. In
addition, these hotel factors may moderate the response effect on review volume. For high-
end, higher-rated and branded hotels, customers would usually expect a higher standard of
service (Xie et al., 2016). Accordingly, customers’ expectation of firms being attentive and
responsiveness is potentially stronger, leading to an increased incentive to engage and
interact with firms via the virtual platform. It is proposed that:
Hypothesis 4: The effects of responsiveness (response volume, speed and length) on future
review volume is positively moderated by hotel characteristics (star, rating, chain).
4. Methods
4.1 Data and sample
This research explains online managerial responses in relation to review volume. Data on
review and response of London hotels is collected from a leading travel site. London is
chosen because it is a globally popular travel destination and a highly competitive market,
which has an extensive number of hotels and reviews/responses that meet sampling and
research needs. Information about all the formal hotels listed on the site at the time of data
collection (including hotel identification, star class, and number of rooms) and the review and
response history – including review date, review title and text, reviewer identification,
response date, and response text – from the first review post of each hotel to the data entries
at the time of data collection (i.e., early 2016) are downloaded. The raw dataset is cleaned for
13
the subsequent empirical analysis. First, hotels with zero reviews are removed (i.e., 26 hotels)
as no observations of review and response are available to study the firm and customer
engagement behaviour. Hotels that are closed at a later stage of the sample period are also
excluded, leading to a loss of 13 more hotels. This is because it is hard to determine the
causes of changes in online posting behaviour of both customers and service providers, which
may bias the results. The final sample contains 1,024 hotels over an about 15-year period
from January 2001 to February 2016. Although not all hotels in the sample appear on the site
at the same time, the cut-off date for the review posts is the end of February 2016.
Table 1 describes the sample. Among the 1,024 hotels, 739 hotels (72.17%) have provided at
least one response in the sample period. High-end hotels are most active in responding to
online reviews (92.7% and 93.6% for the four-star and five-star class respectively). High
customer-rated hotels (rating greater than or equal to 3) also actively engage in online
managerial response, particularly those rated 4 or 4.5. Along the timeline, there is a clear
upward trend in writing online reviews and responses (see Figure 2). The overall response
ratio surges in the year 2009 and in general service providers respond to 45.93% of online
customer reviews.
------------------------------
Insert Table 1 about here
------------------------------
------------------------------
Insert Figure 2 about here
------------------------------
In addition, Figure 3 presents the percentage of customer ratings on a scale of 1 to 5 for each
year over the sample period. Overall, the tone of the collected comments tends to be positive.
In particular, after the year 2009 when the response ratio started to accelerate, there is an
14
increase in the proportion of 5-score reviews while a decrease in that of 1- or 2-score reviews.
Neutral reviews with a rating of 3 and 4 make up a relatively fixed percentage of the reviews
over the years. This may reflect quality improvement in the hotel sector and possible changes
in customer rating behaviour as a result of enhanced online firm–customer communications.
------------------------------
Insert Figure 3 about here
------------------------------
4.2 Variables and models
Data is organised at Hotel–Month level. Following the approach in earlier research (e.g.,
Duan et al., 2008; Gu and Ye, 2014; Xie et al., 2014), customer reviews and management
responses in previous time periods are used to assess the influence on later customers, as the
WOM effects often last for several weeks (Trusov et al., 2009; Xie et al., 2014). The
dependent variable is a hotel’s number of reviews in a calendar month (ReviewVolume), and
the explanatory variables are the three variables of responsiveness (ResponseVolume,
ResponseDays, and ResponseLength) in the previous month. Besides, as discussed in the
previous section, hotels providing quality services may naturally attract more consumers to
stay, potentially leading to a higher volume of reviews, and these hotels are more likely to
provide managerial responses in a professional way, leading to a higher level of
responsiveness. Therefore, five hotel-specific factors—Star, Rating, Chain, Size, and Age—
are included to control for hotel-level heterogeneity (see Table 2 for a detailed description of
variables).3 To control for time effect, a time dummy for each month is also included to
�������������������������3 The value of these variables is at a fixed time point (i.e., the time point of data collection), but these attributes
are not strictly time-invariant (e.g., star class upgrade/downgrade, size expansion, brand acquisition etc.). The
purpose of including these variables in the model is to account for differences among sampled hotels in these
aspects.
15
account for time trend of review and response behaviour that is common to all hotels.
Considering that the data on monthly review and response is clustered at the hotel level, a
multilevel model (two-level model) is adopted, with the review/response at the first level and
the hotel as the second level indicator. The random effects model estimates the group effects
and group level predictors at the same time. Furthermore, there might be time effects and
such time effects may vary across individual hotels. Hence, the time factor (i.e., month) is
included at the group level to allow for random slopes across different hotels. The model is
specified as:
"#$%#&'()*+#,- = 01"#23(42%$#4#22,-51 + 718%9+, + :, + *;,+*,- + #,-
where Responsiveness is tested with three response variables, including ResponseVolume,
Responsedays, and ResponseLength. The dependent variable and the response variables are at
monthly level (taking the logarithmic values), and variables of Responsiveness are one month
lagged; Firmh is a vector of hotel specific factors—Star, Rating, and Chain; ti includes other
hotel factors, Size and Age (taking the logarithmic values), and time dummies; u0h and uht
capture the random effects of hotel h and time t and eht are observation-level residuals.
Multicollinearity is not a concern for the variables of interest given the low VIFs compared to
the common threshold (see Table 2).
------------------------------
Insert Table 2 about here
------------------------------
5. Results
5.1 Model-free evidence
A two-sample t-test is conducted to determine if there is any significant difference in daily
16
review volume (measured as the daily number of reviews each hotel receives during the
period of its presence on the website) between responding and non-responding hotels. Table
3 shows that responding hotels on average receive 0.452 online reviews per day (equivalent
to 13.748 reviews per month and 164.976 reviews per year), while non-responding hotels
have an average daily review number of 0.065 (equivalent to 1.977 per month and 23.725 per
year). The difference between the two groups is statistically and practically significant (p <
0.001). Such significant difference exists between responding and non-responding hotels at
each star class except for five-star hotels (p = 0.137). The model-free evidence suggests that
review volume of responding hotels is significantly larger than non-responding hotels.
------------------------------
Insert Table 3 about here
------------------------------
5.2 Main results
Estimations results for responsiveness are presented in Table 4 Columns 1–3. First, as
expected in the first hypothesis, response volume is positively associated with review volume
(β = 0.092, p < 0.001). It means a 10% increase in the number of responses relates to 0.92%
increase in the number of reviews in the next period. Second, ResponseDays is negatively
associated with review volume (β = -0.033, p < 0.001). For example, a 10% decrease in the
monthly average time intervals (days) between reviews and the associated responses from the
service provider leads to about 0.33% increase in the review volume in the following month.
This suggests that response speed has a positive impact on review volume, which supports
hypothesis number two. Moreover, the third hypothesis of response length is also supported
given the result showing its positive association with review volume (β = 0.023, p = 0.038). It
implies a 10% increase in the word counts of responses is linked with 0.23% increase in the
17
future review volume. These findings together support presumptions about the positive effect
of responsiveness on future review volume.
In addition, the results show that review volume is largely affected by former reviewers in
terms of how they rate the service. As shown in Table 4, the average customer ratings of
hotels established online are positively related to future review volume. Approximately, a 1
score increase in the average ratings may lead to an over 20% increase in the next period’s
review volume. This implies the significance of the crowd effect on customer engagement
behaviour. Besides, a positive relationship between hotel size and review volume and a
negative relationship between hotel age and review volume are detected. Furthermore, the
results show there is no significant effect of star class and chain brand on review volume.
With regard to the moderating effects of hotel specific factors (i.e., Star, Rating, Chain) on
the response effect, the estimation results in column 4–6 of Table 4 show no evidence to
support hypothesis 4, which is in contrast to the prediction.
------------------------------
Insert Table 4 about here
------------------------------
5.3 Additional tests
Several additional tests are conducted to check the robustness of the results. First, it is worth
pointing out that the time dummies are significant after the year 2009. This is in line with the
growing trend in reviewing and responding behaviour starting from that date. To further
check the robustness, all data before the year 2009 is eliminated. The remaining data in the
period 2009–2016 is used to re-estimate the multilevel model. The estimations (Table 5 Panel
A) confirm the main results that response volume is positively associated with review volume
(β = 0.089, p < 0.001) and response days are negatively associated with review volume (β = -
18
0.032, p < 0.001), and there is no evidence showing that hotel factors can moderate response
effects on review volume. However, in contrast to the based result, response length has no
statistically significant influence (β = 0.016, p = 0.149).
------------------------------
Insert Table 5 about here
------------------------------
Next, the data includes some repeat reviewers who have multiple reviews for the same hotel.
Godes and Mayzlin (2009) demonstrate that the effects of firm-generated messages in the
word-of-mouth marketing campaign vary with the degree of customer loyalty. Their findings
suggest that exogenous word-of-mouth created by firms is more impactful and raises
awareness among less loyal customers because they are less informed than loyal customers,
who have already formed strong ties and opinions about the firm. Gu and Ye (2014) also hint
that there might be a self-selection issue among returning customers who are more likely to
write reviews. The information distortion derived from individual preference may affect the
decision-making process (Chaxel & Han, 2018). Therefore, we can exclude the reviews
written by returning customers (i.e., a customer writes more than two reviews of the same
hotel in the sample period) to check the sensitivity of results to customers’ heterogeneous
preference. Panel B of Table 5 shows that the results are robust to measuring one-time
reviewers, except for response length (β = 0.015, p = 0.175). The response volume remains
significantly positive (β = 0.092, p < 0.001), and response days present a negative
relationship (β = -0.034, p < 0.001).
In addition, as presented in Panel C of Table 5, the investigation focuses on the responding
hotels only after they start to respond. A subsample is created only keeping review
observations of responding hotels after the date of each responding hotel’s first managerial
response. The subsample includes 692 hotels, 642,501 customer reviews, and 358,752
19
managerial responses from March 2004 to February 2016. Estimations are very similar to the
baseline results. The indicators of responsiveness retain a significant and positive relationship
with the review volume in the following period (ResponseVolume, β = 0.085, p = 0.001;
ResponseDays, β = -0.029, p < 0.001; ResponseLength, β = 0.019, p = 0.095). These effects
are not moderated by hotel specific factors.
Finally, in the main test, data is organised at the monthly level and the results may be
sensitive to the choice of the time window. To rule out this possibility, the model is estimated
respectively using a weekly and quarterly time window (not reported in the table). Consistent
with the baseline random effect estimations, results show that the number of responses is
positively associated with future review volume. Response days are negatively related to
review volume, suggesting a positive effect of response speed on future review volume. But
there is no evidence supporting the relationship between response length and review volume
and the moderating effects of hotel factors.
6. Discussions and Conclusions
6.1 Summary of findings and theoretical implications
This study examines online firm and customer engagement issue by studying the behavioural
effect of managerial responses on customer reviews. The sampled data presents a fact that
responding firms have a larger number of daily review volume compared to non-responding
hotels. This provides extra evidence to prior studies (e.g., Ye et al., 2010; Proserpio & Zervas,
2017; Chevalier et al., 2017) which discover a positive relationship between providing online
managerial responses and customer review volume. Further, in testing the multilevel random
effect model, it is found that business responsiveness has a strong relation with future review
volume. In particular, the empirical results show a significant and positive influence of
response volume on future review volume, which is in line with the conclusion in Xie et al.
20
(2016). Besides, a novel finding in this research is that response speed is a strong indicator of
firm responsiveness which positively influences customers’ participation in writing
comments. This echoes the significance of timing in the service recovery literature (e.g.,
Davidow, 2003; Homburg & Fürst, 2007; Sparks et al., 2016); but instead of accentuating the
effect on low-satisfaction consumers, this research highlights the promptness of responses to
all potential reviewers. These findings support the first two hypotheses, implying that
responding frequently and quickly can lead to an increase in the number of reviews in the
longer term.
In addition, inconsistent with the hypothesis number three, there is limited evidence showing
the significance of response length in relation to future review volume. Different from the
expectation of an informational role of responses, the empirical evidence suggests that the
possible effect on review engagement is trivial. Besides, no evidence is documented to
support the last hypothesis. Although some hotel specific factors play a role in shaping the
likelihood of customers’ review engagement, they cannot moderate the impact of
responsiveness on review volume. This implies that online responsiveness is critical
notwithstanding the level, type and capability of firms.
Altogether, these findings suggest that customers’ engagement intention and behaviour are
influenced by firms’ engagement in the online conversations. This contributes to the
engagement literature (e.g., Eisingerich et al., 2015; Mathwick & Mosteller, 2017; Pansari &
Kumar, 2017; Van Doorn et al., 2010) by determining that firm engagement is a motivational
driver of customer engagement behaviour. Apart from self-motivation for word-of-mouth
sharing (Berger, 2014), there is a spill over effect of the managerial intervention on reviewing
behaviour of the community members. A business being responsive and active on the social
media can facilitate interactions between customers and firms, which can attract, encourage
21
and stimulate online users, especially potential reviewers, to engage in online reviewing and
communications. This research also contributes to the marketing research in relation to social
media efforts by investigating the effect of online firm-generated messages that has been
understudied in the current literature (Harmeling et al., 2017; Kumar et al., 2016). Prior
studies tend to estimate response effects before and after the policy change rather than the
long-term effect. It remains unclear what the key factors are that affect firm responsiveness
and hence how it exerts an influence on customers’ engagement in writing reviews. Studying
the behavioural effects of firm responsiveness in an online review context suggests that firms’
strategic participation in online communications can potentially create leading influence and
draw wider attention, which makes it an effective tool to enhance online popularity and social
influence.
6.2 Implications for practice
These discussions clearly show that firms’ online responsiveness can stimulate customer
engagement behaviour in eWOM communications. The business’s strategic and voluntary
exposure on online social sites can help gain customers’ attention, expand the consumer
network, manage customers, and enhance social influence and online popularity, all
potentially leading to favourable outcomes. This requires firms to make strategic changes
with “committing to long-term paths or trajectories of competence development” (Teece et al.,
1997, p. 529). For firms that have not established an online presence in the network,
providing managerial responses would be an option to kick-start engagement in online firm-
customer communications and active management of their social media presence. For firms
that have adopted social media to implement marketing activities, it is important to keep the
engagement and communication as a consistent practice. Businesses should respond in a
faster and frequent way to make sure the managerial effort is manifest to customers.
22
Especially when a firm receives a large number of reviews in a certain period, it is important
to compete with the review update speed. Responding quickly and frequently increases the
possibility of responses being displayed on the first few pages and thus being easier for
review readers and potential reviewers to see, leading to an enhanced power in influencing
the propensity for customer engagement. The continuous and positive impact of firm
responsiveness creates strategic value for managing customers and potentially for financial
outcomes.
6.3 Limitations and future research
A few limitations should be acknowledged. First, this study focuses on online popularity as
demonstrated by the number of customer reviews. It does not consider offline popularity,
such as the actual number of visitors, and its potential influence on the review volume. Future
research may extend this study by examining the relationship between offline and online
popularity and the possible impact of managerial response on sales/revenue generation.
Second, the included control variables of hotel characteristics are not exhaustive. Additional
variables such as price, location, and unobservable attributes (e.g., improvements to hotels’
managerial expertise and service quality) can be added to the model to assess the offline
popularity and dynamics. Third, the research setting to investigate the business social media
presence and activeness in this study is an online community-based review platform. This is a
third-party organised communication channel, which may present some policy-related issues
that impede or affect how firms engage. Furthermore, the review-response communication is
less firm-initiated. It would be interesting to examine the interplay between firm engagement
and customer engagement behaviour by using “firm-initiated marketing communication in its
official social media pages” (Kumar et al., 2016, p. 7), given that the corporate resources
allocated to managing the channels are different (Ashley & Tuten, 2015).
23
References
Ashley, C., & Tuten, T. (2015). Creative strategies in social media marketing: An exploratory
study of branded social content and consumer engagement. Psychology &
Marketing, 32(1), 15-27.
Barrales-Molina, V., Martínez-López, F. J., & Gázquez-Abad, J. C. (2014). Dynamic
marketing capabilities: Toward an integrative framework. International Journal of
Management Reviews, 16(4), 397-416.
Berger, J. (2014). Word of mouth and interpersonal communication: A review and directions
for future research. Journal of Consumer Psychology, 24(4), 586-607.
Bolton, R. N. (2011). Comment: Customer engagement: Opportunities and challenges for
organizations. Journal of Service Research, 14(3), 272-274.
Buechel, E. C., & Berger, J. (2018). Microblogging and the value of undirected
communication. Journal of Consumer Psychology, 28(1), 40-55.
Chang, Y. T., Yu, H., & Lu, H. P. (2015). Persuasive messages, popularity cohesion, and
message diffusion in social media marketing. Journal of Business Research, 68(4), 777-
782.
Chaxel, A. S., & Han, Y. (2018). Benefiting from disagreement: Counterarguing reduces
prechoice bias in information evaluation. Journal of Consumer Psychology, 28(1), 115-
129.
Cheng, Y. H., & Ho, H. Y. (2015). Social influence's impact on reader perceptions of online
reviews. Journal of Business Research, 68(4), 883-887.
Chevalier, J. A., Dover, Y., & Mayzlin, D. (2017). Channels of impact: User reviews when
quality is dynamic and managers respond. National Bureau of Economic Research,
Working Paper, No. w23299.
Davidow, M. (2003). Organizational responses to customer complaints: What works and
what doesn’t. Journal of Service Research, 5(3), 225-250.
24
de Vries, L., Gensler, S., & Leeflang, P. S. (2012). Popularity of brand posts on brand fan
pages: An investigation of the effects of social media marketing. Journal of Interactive
Marketing, 26(2), 83-91.
de Vries, L., Gensler, S., & Leeflang, P. S. (2017). Effects of traditional advertising and
social messages on brand-building metrics and customer acquisition. Journal of
Marketing, 81(5), 1-15.
Dong, B., & Sivakumar, K. (2017). Customer participation in services: Domain, scope, and
boundaries. Journal of the Academy of Marketing Science, 45(6), 944-965.
Duan, W., Gu, B., & Whinston, A. B. (2008). Do online reviews matter?—An empirical
investigation of panel data. Decision Support Systems, 45(4), 1007-1016.
Eisingerich, A. B., Chun, H. H., Liu, Y., Jia, H. M., & Bell, S. J. (2015). Why recommend a
brand face-to-face but not on Facebook? How word-of-mouth on online social sites
differs from traditional word-of-mouth. Journal of Consumer Psychology, 25(1), 120-
128.
Enz, C. A., & Grover, R. A. (1992). The importance of top management visibility for service-
based professionals. Journal of Managerial Issues, 414-423.
Evans, D. (2010). Social media marketing: The next generation of business engagement. John
Wiley & Sons.
Fang, X., Hu, P. J. H., Li, Z., & Tsai, W. (2013). Predicting adoption probabilities in social
networks. Information Systems Research, 24(1), 128-145.
Felix, R., Rauschnabel, P. A., & Hinsch, C. (2017). Elements of strategic social media
marketing: A holistic framework. Journal of Business Research, 70, 118-126.
Friedkin, N. E. (1998). A Structural Theory of Social Influence. Cambridge University
Press.�
Gensler, S., Völckner, F., Liu-Thompkins, Y., & Wiertz, C. (2013). Managing brands in the
social media environment. Journal of Interactive Marketing, 27(4), 242-256.
25
Godes, D., & Mayzlin, D. (2009). Firm-created word-of-mouth communication: Evidence
from a field test. Marketing Science, 28(4), 721-739.
Godey, B., Manthiou, A., Pederzoli, D., Rokka, J., Aiello, G., Donvito, R., & Singh, R.
(2016). Social media marketing efforts of luxury brands: Influence on brand equity and
consumer behavior. Journal of Business Research, 69(12), 5833-5841.
Goes, P. B., Lin, M., & Au Yeung, C. M. (2014). ‘Popularity effect’ in user-generated
content: Evidence from online product reviews. Information Systems Research, 25(2),
222-238.
Goh, K.Y., Heng, C.S., & Lin, Z. (2013). Social media brand community and consumer
behavior: Quantifying the relative impact of user-and marketer-generated content.
Information Systems Research, 24(1), 88-107.
Gong, S., Zhang, J., Zhao, P., & Jiang, X. (2017). Tweeting as a marketing tool–field
experiment in the TV industry. Journal of Marketing Research, LIV, 833-850.
Gu, B., & Ye, Q. (2014). First step in social media: Measuring the influence of online
management responses on customer satisfaction. Production and Operations
Management, 23(4), 570-582.
Harmeling, C. M., Moffett, J. W., Arnold, M. J., & Carlson, B. D. (2017). Toward a theory of
customer engagement marketing. Journal of the Academy of Marketing Science, 45(3),
312-335.
Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-of-mouth via consumer-opinion platforms: What motivates consumers to articulate
themselves on the Internet?. Journal of Interactive Marketing, 18(1), 38-52.
Higgins, E. T., & Scholer, A. A. (2009). Engaging the consumer: The science and art of the
value creation process. Journal of Consumer Psychology, 19(2), 100-114.
Hollebeek, L. D., Glynn, M. S., & Brodie, R. J. (2014). Consumer brand engagement in
social media: Conceptualization, scale development and validation. Journal of
Interactive Marketing, 28(2), 149-165.
26
Homburg, C., & Fürst, A. (2007). See no evil, hear no evil, speak no evil: A study of
defensive organizational behavior towards customer complaints. Journal of the Academy
of Marketing Science, 35(4), 523-536.
Homburg, C., Ehm, L., & Artz, M. (2015). Measuring and managing consumer sentiment in
an online community environment. Journal of Marketing Research, LII, 629-641.
Iyengar, R., Van den Bulte, C., & Valente, T.W. (2011). Opinion leadership and social
contagion in new product diffusion. Marketing Science, 30(2), 195-212.
Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and
opportunities of social media. Business Horizons, 53(1), 59–68.
Kim, A. J., & Ko, E. (2012). Do social media marketing activities enhance customer equity?
An empirical study of luxury fashion brand. Journal of Business Research, 65(10), 1480-
1486.
King, R. A., Racherla, P., & Bush, V. D. (2014). What we know and don't know about online
word-of-mouth: A review and synthesis of the literature. Journal of Interactive
Marketing, 28(3), 167-183.
Kozinets, R.V. (2002). The field behind the screen: Using netnography for marketing
research in online communities. Journal of Marketing Research, 39(1), 61-72.
Kozinets, R.V., De Valck, K., Wojnicki, A.C., & Wilner, S.J. (2010). Networked narratives:
Understanding word-of-mouth marketing in online communities. Journal of Marketing,
74(2), 71-89.
Kumar, A., Bezawada, R., Rishika, R., Janakiraman, R., & Kannan, P. K. (2016). From social
to sale: The effects of firm-generated content in social media on customer behavior.
Journal of Marketing, 80, 7-25.
Kumar, V., Bhaskaran, V., Mirchandani, R., & Shah, M. (2013). Practice prize winner—
creating a measurable social media marketing strategy: increasing the value and ROI of
intangibles and tangibles for hokey pokey. Marketing Science, 32(2), 194-212.
27
Kurt, D., Inman, J.J., & Argo, J.J. 2011. The influence of friends on consumer spending: The
role of agency-communion orientation and self-monitoring. Journal of Marketing
Research, 48(4), 741-754.
Lamberton, C., & Stephen, A. T. (2016). A thematic exploration of digital, social media, and
mobile marketing: research evolution from 2000 to 2015 and an agenda for future
inquiry. Journal of Marketing, 80(6), 146-172.
Lee, Y.J., Hosanagar, K., & Tan, Y. (2015). Do I follow my friends or the crowd?
Information cascades in online movie ratings. Management Science, 61(9), 2241-2258.
Mathwick, C., & Mosteller, J. (2017). Online reviewer engagement: A typology based on
reviewer motivations. Journal of Service Research, 20(2), 204-218.
O’Reilly, T. (2005). What is Web 2.0? Design patterns and business models for the next
generation of software,
http://www.oreillynet.com/pub/a/oreilly/tim/news/2005/09/30/what-is-web-20.html
Pansari, A., & Kumar, V. (2017). Customer engagement: The construct, antecedents, and
consequences. Journal of the Academy of Marketing Science, 45(3), 294-311.
Pavlou, P. A., & Dimoka, A. (2006). The nature and role of feedback text comments in online
marketplaces: Implications for trust building, price premiums, and seller
differentiation. Information Systems Research, 17(4), 392-414.
Proserpio, D., & Zervas, G. (2017). Online reputation management: Estimating the impact of
management responses on consumer reviews. Marketing Science, 36(5), 645-665.
Schniederjans, D., Cao, E. S., & Schniederjans, M. (2013). Enhancing financial performance
with social media: An impression management perspective. Decision Support
Systems, 55(4), 911-918.
Shoham, M., Moldovan, S., & Steinhart, Y. (2017). Positively useless: Irrelevant negative
information enhances positive impressions. Journal of Consumer Psychology, 27(2),
147-159.
28
Shriver, S. K., Nair, H. S., & Hofstetter, R. (2013). Social ties and user-generated content:
Evidence from an online social network. Management Science, 59(6), 1425-1443.
Sparks, B. A., So, K. K. F., & Bradley, G. L. (2016). Responding to negative online reviews:
The effects of hotel responses on customer inferences of trust and concern. Tourism
Management, 53, 74-85.
Sridhar, S., & Srinivasan, R. (2012). Social influence effects in online product
ratings. Journal of Marketing, 76(5), 70-88.
Srinivasan, S., Vanhuele, M., & Pauwels, K. (2010). Mind-set metrics in market response
models: An integrative approach. Journal of Marketing Research, 47(4), 672-684.
Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic
management. Strategic Management Journal, 509-533.
Tirunillai, S., & Tellis, G. J. (2012). Does chatter really matter? Dynamics of user-generated
content and stock performance. Marketing Science, 31(2), 198-215.
Trusov, M., Bucklin, R. E., & Pauwels, K. (2009). Effects of word-of-mouth versus
traditional marketing: Findings from an internet social networking site. Journal of Marketing,
73(5), 90-102.
van Doorn, J., Lemon, K. N., Mittal, V., Nass, S., Pick, D., Pirner, P., & Verhoef, P. C.
(2010). Customer engagement behavior: Theoretical foundations and research
directions. Journal of Service Research, 13(3), 253-266.
Verhoef, P. C., Reinartz, W. J., & Krafft, M. (2010). Customer engagement as a new
perspective in customer management. Journal of Service Research, 13(3), 247-252.
Wang, J., Aribarg, A., & Atchadé, Y.F. (2013). Modeling choice interdependence in a social
network. Marketing Science, 32(6), 977-997.
Wang, Y., & Chaudhry, A. (2018). When and How Managers' Responses to Online Reviews
Affect Subsequent Reviews. Journal of Marketing Research, 55(2), 163-177.
29
Wang, Z., & Kim, H. G. (2017). Can social media marketing improve customer relationship
capabilities and firm performance? Dynamic capability perspective. Journal of
Interactive Marketing, 39, 15-26.
Xie, K. L., Zhang, Z., & Zhang, Z. (2014). The business value of online consumer reviews
and management response to hotel performance. International Journal of Hospitality
Management, 43, 1-12.
Xie, K. L., Zhang, Z., Zhang, Z., Singh A., & Lee, S. K. (2016). Effects of managerial
response on consumer eWOM and hotel performance: Evidence from
TripAdvisor. International Journal of Contemporary Hospitality Management, 28(9),
2013-2034.
Yadav, M. S., De Valck, K., Hennig-Thurau, T., Hoffman, D. L., & Spann, M. (2013). Social
commerce: A contingency framework for assessing marketing potential. Journal of
Interactive Marketing, 27(4), 311-323.
Ye, Q., Gu, B., & Chen, W. (2010). Measuring the influence of managerial responses on
subsequent online customer reviews–a natural experiment of two online travel agencies,
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1639683
Zhang, M., Jansen, B.J., & Chowdhury, A. (2011). Business engagement on Twitter: A path
analysis. Electronic Markets, 21(3), 161-175.
Zhang, X., Li, S., Burke, R. R., & Leykin, A. (2014). An examination of social influence on
shopper behavior using video tracking data. Journal of Marketing, 78(5), 24-41.
�
30
Figures
Figure 1. Research framework
Note: The year of review and the year of its associated response may be different. Yearly distribution of
response ratio is based on the year when the reviews were posted. The drop of review/response number in 2016
is due to data availability. The data was collected in March 2016 and refined to the period before the end of
February 2016.
Figure 2. Number of reviews and responses
�
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016Number of Reviews 1 132 1,261 3,971 5,238 8,050 11,172 13,509 23,208 32,939 56,310 94,233 134,113165,813200,952 36,349
Number of Responses 0 0 0 8 11 65 108 299 2,696 5,969 16,992 38,269 64,594 91,223 122,097 19,286
Response Ratio 0 0 0.95% 0.73% 1.15% 1.40% 1.50% 3.28% 13.60% 19.30% 32.32% 41.49% 48.20% 55.05% 60.16% 47.22%
0
0.1
0.2
0.3
0.4
0.5
0.6
0
20000
40000
60000
80000
100000
120000
140000
160000
180000
200000
Res
pons
eRat
io
No.
of R
evie
ws/
Res
pons
es
Year
Review Volume
Response Volume
Response Speed
H1
H2
Response Length
H3
Bus
ines
s Res
pons
iven
ess
Online Popularity
Hotel Characteristics
Star, Rating, Chain
Control: Size, Age H4
31
Figure 3. Distribution of ratings over time
�
16.54 14.47 15.01 12.81 14.24 13.24 12.88 10.22 9.99 7.69 5.66 4.89 5.15 4.86 3.94 6.09
10.53 14.39 11.9911.3 11.75 11.59 11.25
7.93 8.026.87 6.29 5.88 5.92 5.53 4.86
6.37
9.7712.09 13.46
13.44 13.54 14.09 14.0816.24 16.79
15.9316.07 15.55 14.79 14.55 12.92
15.08
30.8330.51 31.69
30.78 31.24 30.82 30.9531.51 32.23
34.09 34.94 34.8 33.63 32.2731.74
33.3
32.33 28.54 27.85 31.67 29.23 30.26 30.84 34.09 32.97 35.42 37.04 38.87 40.51 42.8 46.5439.15
0
10
20
30
40
50
60
70
80
90
100
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Total
Perc
enta
ge
Year
5
4
3
2
1
32
Tables
Table 1. Descriptive statistics
Star class Number
of hotels
Number of
responding hotels Rating
Number
of hotels
Number of
responding hotels
5 110 103 (93.64%) 5 44 29 (65.91%)
4/4.5 274 254 (92.70%) 4/4.5 506 445 (87.94%)
3/3.5 361 282 (78.12%) 3/3.5 289 199 (68.86%)
2/2.5 133 67 (50.38%) 2/2.5 154 60 (38.96%)
0/1/1.5 146 33 (22.60%) 1/1.5 31 6 (19.35%)
Total 1024 739 (72.17%) Total 1024 739 (72.17%)
Note: The null value of star class is due to unavailability of this information on the review
website. Rating is the overall customer-rated score, which is round to the nearest .5.
33
Table 2. Summary statistics
Variables Description N Mean SD VIF
ReviewVolumeht The logarithm of hotel h’s number of
reviews in period t
76,329 1.567 1.215 2.630
ResponseVolumeht-1 The logarithm of hotel h’s number of
responses in period t-1
23,847 2.029 1.223 1.890
ResponseDaysht-1 The logarithm of hotel h’s average number
of days between responses and the
associated reviews in period t-1
23,845 2.131 1.239 1.130
ResponseLengthht-1 The logarithm of hotel h’s average word
count of responses in period t-1
23,582 4.385 0.496 1.050
Starh Hotel h’s star class on a five-star scale 76,329 3.323 1.133 1.550
Ratingh Hotel h’s overall average customer review
ratings on a scale of 1 to 5
76,329 3.646 0.796 1.410
Chainh An indicator variable, which takes the value
of 1 if hotel h is a chain hotel and takes the
value of 0 if hotel h is an independent hotel
76,329 0.265 0.441 1.310
Sizeh The logarithm of hotel h’s number of rooms 75,849 4.283 1.037 1.910
Ageh The duration of presence on the website,
measured by the logarithm of days from the
date of hotel h’s first review to the cut-off
date
76,328 8.197 0.451 1.120
Note: ReviewVolume, ResponseVolume, ResponseDays, ResponseLength, Size and Age take
logarithmic values. ResponseVolume, ResponseDays, ResponseLength are one month lagged.
34
Table 3. T-test for review volume between responding and non-responding hotels
Responseihv-1=0 Responseihv-1=1 t-test
Variable Conditions M SD M SD t-value
DailyReviewVolumeh 0.065 0.120 0.452 0.575 -11.250***
Star = 0/1/1.5 0.033 0.117 0.255 0.287 -6.620***
Star = 2/2.5 0.065 0.055 0.242 0.353 -4.037***
Star = 3/3.5 0.096 0.127 0.358 0.534 -4.314***
Star = 4/4.5 0.079 0.133 0.600 0.635 -3.656***
Star = 5 0.193 0.263 0.542 0.612 -1.497
Note: Response is an indicator variable, demonstrating whether a hotel provides
managerial responses, which takes the value of 1 if hotel h has provided at least one
online response in the sample period. DailyReviewVolume is measured as the daily
number of reviews in the duration of hotel presence on the site. Star is the hotel star
class. Rating is the overall average customer review rating. * p < 0.1, ** p < 0.05, *** p
< 0.01
35
Table 4. Effects of responsiveness on review volume
ReviewVolumeht (1) (2) (3) (4) (5) (6)
Fixed effects
ResponseVolumeht-1 0.092*** 0.071
(0.006) (0.044)
ResponseDaysht-1 -0.033*** -0.002
(0.004) (0.027)
ResponseLengthht-1 0.023** -0.006
(0.011) (0.075)
Starh -0.049 -0.045 -0.050 -0.053 -0.022 0.034
(0.045) (0.049) (0.050) (0.045) (0.052) (0.094)
Ratingh 0.238*** 0.239*** 0.265*** 0.240*** 0.244*** 0.126
(0.065) (0.072) (0.073) (0.066) (0.075) (0.130)
Chainh -0.157* -0.120 -0.127 -0.157* -0.131 0.183
(0.084) (0.092) (0.092) (0.084) (0.094) (0.143)
Sizeh 0.546*** 0.546*** 0.556*** 0.547*** 0.545*** 0.559***
(0.044) (0.049) (0.049) (0.044) (0.049) (0.049)
Ageh -0.147*** -0.159*** -0.157*** -0.148*** -0.160*** -0.159***
(0.042) (0.048) (0.047) (0.042) (0.048) (0.047)
Responsivenessht-1×Starh 0.012 -0.007 -0.019
(0.008) (0.005) (0.016)
Responsivenessht-1×Ratingh -0.006 -0.001 0.030
(0.012) (0.008) (0.023)
Responsivenessht-1×Chainh 0.006 0.003 -0.070***
(0.012) (0.008) (0.024)
Time dummy Yes Yes Yes Yes Yes Yes
36
Intercept -1.949*** -1.717*** -2.186*** -1.931*** -1.790*** -2.026***
(0.460) (0.516) (0.521) (0.456) (0.519) (0.615)
Random effect variances
Hotel level 0.666*** 0.833*** 0.823*** 0.661*** 0.836*** 0.816***
(0.103) (0.118) (0.118) (0.104) (0.118) (0.117)
Month time effect 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Monthly review/response level 0.165*** 0.166*** 0.167*** 0.165*** 0.166*** 0.167***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
N 23676 23674 23413 23676 23674 23413
Log-likelihood -14268.984 -14461.145 -14319.342 -14266.224 -14458.027 -14310.178
Note: The three independent variables of responsiveness are one month lagged. ReviewVolume, ResponseVolume,
ResponseDays, ResponseLength, Size and Age take logarithmic values. The variable of responsiveness in the
interaction terms for column 4, 5, 6 is ResponseVolume, ResponseDays, and ResponseLength respectively. The
multilevel models present maximum likelihood estimations. All estimations have robust error terms clustered at the
hotel level. Standard errors in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01
37
Table 5. Robustness checks
ReviewVolumeht Panel A: After the year 2009 Panel B: One-time reviewer Panel C: After responding
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Fixed effects
ResponseVolumeht-1 0.089*** 0.092*** 0.085***
(0.006) (0.006) (0.006)
ResponseDaysht-1 -0.032*** -0.034*** -0.029***
(0.004) (0.004) (0.004)
ResponseLengthht-1 0.016 0.015 0.019*
(0.011) (0.011) (0.011)
Starh -0.050 -0.042 -0.046 -0.052 -0.046 -0.049 -0.103** -0.102** -0.102**
(0.048) (0.053) (0.053) (0.044) (0.048) (0.049) (0.046) (0.050) (0.051)
Ratingh 0.289*** 0.299*** 0.318*** 0.244*** 0.241*** 0.252*** 0.383*** 0.411*** 0.415***
(0.070) (0.077) (0.077) (0.067) (0.074) (0.076) (0.061) (0.067) (0.067)
Chainh -0.139 -0.096 -0.105 -0.137* -0.101 -0.098 -0.091 -0.048 -0.050
(0.086) (0.094) (0.093) (0.081) (0.089) (0.089) (0.077) (0.084) (0.084)
Sizeh 0.570*** 0.573*** 0.582*** 0.537*** 0.537*** 0.540*** 0.585*** 0.595*** 0.597***
(0.046) (0.051) (0.050) (0.042) (0.047) (0.048) (0.042) (0.046) (0.046)
38
Ageh -0.140*** -0.150*** -0.149*** -0.153*** -0.169*** -0.166*** 0.127*** 0.173*** 0.165***
(0.043) (0.048) (0.048) (0.041) (0.046) (0.047) (0.045) (0.049) (0.050)
Time dummy Yes Yes Yes Yes Yes Yes Yes Yes Yes
Intercept -1.540*** -1.489*** -1.705*** -1.796*** -1.501*** -1.869*** -3.412*** -3.788*** -3.907***
(0.481) (0.534) (0.532) (0.453) (0.510) (0.523) (0.400) (0.436) (0.448)
Random effect variances
Hotel level 0.720*** 0.888*** 0.877*** 0.605*** 0.762*** 0.780*** 0.592*** 0.717*** 0.718***
(0.123) (0.135) (0.134) (0.099) (0.112) (0.117) (0.078) (0.088) (0.089)
Month time effect 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Monthly review/response level 0.163*** 0.164*** 0.164*** 0.171*** 0.172*** 0.172*** 0.159*** 0.160*** 0.160***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
N 23317 23315 23059 23556 23553 23285 22857 22854 22644
Log-likelihood -13894.957 -14074.261 -13937.817 -14519.572 -14699.255 -14556.196 -13287.823 -13456.337 -13358.066
Note: The three independent variables of responsiveness are one month lagged. ReviewVolume, ResponseVolume, ResponseDays, ResponseLength, Size and
Age take logarithmic values. The variable of responsiveness in the interaction terms for column 4, 5, 6 is ResponseVolume, ResponseDays, and
ResponseLength respectively. The multilevel models present maximum likelihood estimations. All estimations have robust error terms clustered at the hotel
level. Standard errors in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01